Infleqtion
This is the archived full research. Read the story
A maker of atom-based clocks and sensors for governments, priced as if its quantum computer were already a winner.
The scores
Growth3/4 →
Up 57% over nine months with $36.1M of work scheduled for 2027, but part of 2026 is revenue moved from earlier years.
Business quality1/4 →
A gross loss in Q3 2026, one customer at 81%, and rivals in both computing and clocks with outside endorsement or larger scale.
Financial position3/4 →
$560.0M of cash and securities and no debt fund about five years of losses, but nothing yet pays its own way.
Execution record2/4 ↓
Met the logical-qubit target and won awards, but revised three periods of revenue, filed late and missed the quarter's estimate.
Per share capture1/4 →
Stock pay about the size of revenue, options and warrants worth about 16% more shares, and insiders sold, not bought.
Potential3/4 →
Governments spend about $1.0B a year on quantum research and DARPA pays up to $300M a finalist, but nobody counts the share companies win.
Risk profile1/4 ↓
One customer at 81% of revenue, weak controls, and a computer an outside judge passed over.
Valuation1/4 ↑
About 29 times sales for a business losing more than its revenue, with about $1.1B resting on the computer.
Evidence quality2/4 →
Filed revenue and cash are solid; the sensing and computing split, the logical-qubit result and the pipeline rest on management's words.
If it works 1.7× · if it fails you keep about 36%.
The Verdict
We rate Infleqtion Unattractive. The business is weak: nearly all its sales are development work for governments, one customer paid 81% of the latest quarter, it spends more than it sells, and it has already had to correct its accounts once. The shares are priced for perfection: after valuing the sensor work generously, the price still puts over a billion dollars on a quantum computer that sells about $15M a year and that a Pentagon research agency passed over for a rival using the same atoms. As a bet it is priced for success. A selection by that agency would argue against us.
The long caseGovernments keep paying for atom-based clocks and sensors while Infleqtion's computer reaches enough reliable qubits to win large national orders.
How we weighed it
The price against our value
Against it
No fair value is drawn: losses run every quarter. The price pays far more for the computer than its sales or any outside selection support.
The business
Against it
Weak
A weak business: development work at a gross loss, one customer most of revenue, and rivals with more scale or an outside endorsement.
Growth against what the price assumes
In its favour
Revenue grew fast this year with work already scheduled for next, though part of the growth is revenue moved from earlier years.
What could go wrong
Against it
One customer, weak accounting controls and a revision already made, and a computer roadmap whose payoff waits for later milestones.
Against competitors
Against it
A neutral-atom rival won the government benchmark Infleqtion did not, and IonQ now sells optical clocks too.
How its group is behaving
Cuts both ways
Governments fund quantum work steadily, but the listed group is priced on excitement and fell hard this autumn.
The latest news and report
Cuts both ways
The last report met the qubit target and kept the outlook but missed the revenue estimate, and new awards keep coming.
The evidence under this - 6 claims
A government customer funding several programmes is harder to lose at once than a commercial buyer, though the 10-Q does not name it.
The company attributes part of the gross loss to stock pay and the timing of a NASA contract's costs, so the gross loss is partly not recurring.
The company judged the adjustments not material to the earlier periods.
DARPA opened a new Stage A solicitation due 30 November 2026, so a later entry remains possible.
The vendor's bridge reads the $640.4M of total assets as preferred stock and reaches about $1,959M; the filed balance sheet shows no preferred stock.
Pure-play rivals trade at hundreds of times sales, so the market may value a computing roadmap that is on schedule far above any multiple of today's sales.
The company
Infleqtion, based in Colorado, cools atoms with lasers until they barely move and then uses them. In its clocks the atoms keep time finely enough to replace satellite timing when signals are jammed. In its sensors they detect radio signals or tiny changes in gravity. In its computer, Sqale, each atom held in a grid of laser beams is a qubit, the unit a quantum computer calculates with.
It joined the stock market in February 2026 by merging with a listed shell company.
The evidence under this - 2 claims
These are the company's own measurements, and no outside body has published a check of them.
How it makes money
In the third quarter of 2026, $6.5M of its $9.1M of sales was services, mostly government development contracts paid as the work progresses, and $2.7M was hardware. Management says about two thirds of sales come from sensors and one third from the computer.
One customer paid 81% of the quarter. The work cost more to deliver than it brought in, a loss of $1.1M before any research or overheads.
The evidence under this - 4 claims
A government customer funding several programmes is harder to lose at once than a commercial buyer, though the 10-Q does not name it.
The 10-Q reports one segment and no split by product line, so the mix cannot be checked.
The company attributes part of the gross loss to stock pay and the timing of a NASA contract's costs, so the gross loss is partly not recurring.
Where the money comes from, stream by stream
| Stream | Share of sales (Q3 2026) | How it earns it |
|---|---|---|
| Services | 71% | Development contracts for governments, recognised as the work progresses, plus software and cloud access. |
| Products | 29% | Delivered hardware: clocks, sensors and computer components. |
Who runs it
Matthew Kinsella, a former venture investor on the board since 2018, has run the company since 2024; Pranav Gokhale, whose software start-up Infleqtion bought, is technology chief. Officers and directors owned 6.0% after the merger.
The chief executive was paid $465,474 in 2025, with a bonus at the board's discretion. In May 2026 he sold 769,954 shares near $16 to $18, and Maverick Capital, a large holder with a board seat, sold about 16.5M, none under a pre-set plan.
The evidence under this - 4 claims
Maverick has since sold most of its holding, so the table overstates it today.
A fund leaving a venture position after a listing is selling for its own investors, and the chief executive kept most of his stake.
Why it wins, and who could beat it
Why customers come back
Governments that fund a sensor from development to deployment rarely switch supplier midway, because a new design means new testing and delay. NASA has now put $40M into Infleqtion's gravity sensor, and $36.1M of contracted work is scheduled for 2027.
That edge is a contractor's, not a product company's: no repeat product order outside government development has been disclosed.
The evidence under this - 4 claims
Both are development contracts that pay as work is done, not product sales.
Government contracts can be cut or ended for convenience, so obligations called non-cancelable remain exposed to budgets.
A government customer funding several programmes is harder to lose at once than a commercial buyer, though the 10-Q does not name it.
Research spending is not all procurement; most of it goes to laboratories and universities rather than to companies selling products.
The threat
In computers, IonQ, IBM and Rigetti sell research machines to the same governments and laboratories, and two private companies, Atom Computing and QuEra, use the same atom approach. A Pentagon research agency, DARPA, has picked both of them for its programme to find a useful quantum computer, but not Infleqtion. In clocks, IonQ bought an optical-clock maker in 2025.
The evidence under this - 4 claims
DARPA opened a new Stage A solicitation due 30 November 2026, so a later entry remains possible.
The rivals: IONQ, IONQ, IBM, RGTI, QBTS, QNT, SITM, and 3 more that are private, in-house or listed abroad.
Who else does this, rival by rival
| Ticker | Rival | Where it competes | How far along | Priced against it |
|---|---|---|---|---|
| IONQ | IonQ's trapped-ion computers and its Vector Atomic optical clocksPriced against it: Yes | Quantum computing: Sqale systems and cloud access | On sale | Yes |
| IonQ's Vector Atomic optical clocks | Quantum sensing, timing and software contracts | On sale | ||
| Private | Atom Computing and QuEra, private neutral-atom computer makersPriced against it: Private | Quantum computing: Sqale systems and cloud access | On sale | Private |
| IBM · RGTI | IBM's and Rigetti's superconducting computersPriced against it: Yes | Quantum computing: Sqale systems and cloud access | On sale | Yes |
| Private | PsiQuantum, a private photonic computer developerPriced against it: Private | Quantum computing: Sqale systems and cloud access | In research | Private |
| Not listed | Classical antennas, receivers and atomic clocks from defence suppliersPriced against it: Not listed | Quantum sensing, timing and software contracts | On sale | Not listed |
| QBTS | D-Wave QuantumPriced against it: Yes | - | On sale | Yes |
| QNT | QuantinuumPriced against it: Yes | - | On sale | Yes |
| SITM | SiTimePriced against it: One segment only | - | On sale | One segment only |
How big it could get
How big this could get
The US government spends about $1.0B a year on quantum research, and Infleqtion's last year of sales is about 4% of that. Nobody publishes how much of that money goes to companies rather than to laboratories, so its real market cannot be counted. DARPA pays up to $300M to each computer maker in its final stage; management expects the computer to grow meaningfully only after 2028.
The evidence under this - 4 claims
Research spending is not all procurement; most of it goes to laboratories and universities rather than to companies selling products.
DARPA opened a new Stage A solicitation due 30 November 2026, so a later entry remains possible.
A quarter of nine-month revenue came from the UK, Australia and other buyers, and federal research spending is not the only budget that buys sensors.
The 10-Q reports one segment and no split by product line, so the mix cannot be checked.
The results
What growth costs
Growth has not yet paid for itself. In the third quarter the work cost more than it earned, and research, sales and overheads each ran above sales. Today it looks like a laboratory selling its work rather than a product company gaining scale.
The evidence under this - 3 claims
The company attributes part of the gross loss to stock pay and the timing of a NASA contract's costs, so the gross loss is partly not recurring.
Much of 2026's stock pay is awards that vested when the company listed, so the quarterly rate may fall.
Go-public costs do not recur, and interest on the cash covers part of the loss below the operating line.
The last report (16 Nov)
Quarter ended September 2026, filed figures.
In the third quarter of 2026, sales rose 25% from a year earlier to $9.1M, but fell by a third from the second quarter and came in below the $11.0M analysts expected. The work cost more than it earned, and the loss before interest and tax was $34.5M.
The company kept its outlook of at least $45.1M for the year and said its computer had reached 30 reliable qubits, its target for 2026.
What it did to the thesisIt changed little. The computer stayed on its schedule and the year's outlook held, but sales fell from the quarter before, one customer paid most of them, and the work still cost more than it earned.
The quarter in detail
Sales growth from the previous quarter, filed figures. The company has restated its own history, so a quarter either side of the restatement may describe a different business; n/c marks one that does not compare.
Earnings per share $-0.09 against $-0.05 expected; $-0.09 as filed, with no adjustments.
The evidence under this - 5 claims
Third-quarter revenue fell 32% from the second quarter and came in below the $11.0M analysts expected.
The company attributes part of the gross loss to stock pay and the timing of a NASA contract's costs, so the gross loss is partly not recurring.
At least $45.1M needs about $12.5M in the fourth quarter, above any quarter so far but the second.
No outside body has verified the result, and logical-qubit counts are not measured the same way across companies.
Three analysts make the 2026 figure, and the share of buys fell from 100% of five in July.
What management did with the money
The $560.0M raised by going public sits in short-term bonds earning interest, $5.3M in the third quarter. Pay in shares was $31.3M over nine months, about the size of sales. The company also holds a stake in another quantum firm, Oratomic, valued at $12.7M.
Insiders in 2026Q3: 1 purchases and 10 sales, from their Form 4 filings.
The evidence under this - 4 claims
The accrued-liabilities rise includes tax withheld on employees' shares that must be paid out, so the underlying cash use is larger.
Much of 2026's stock pay is awards that vested when the company listed, so the quarterly rate may fall.
The price, and where we disagree
In short. Priced for perfection. After valuing the sensor work generously, the price still pays over a billion dollars for a computer selling about $15M a year.
The shares are 62.9% below their April high and below the $10 that big investors paid when it went public. About 13% of the shares are sold short.
The next test is the fourth-quarter report, due around 30 March 2027: whether 2026 sales reached $45.1M and how much of 2027 depends on one customer.
The evidence under this - 5 claims
The vendor's float of 53.8M shares, which would make it 56%, is far below the shares outstanding and is not used.
At least $45.1M needs about $12.5M in the fourth quarter, above any quarter so far but the second.
Government contracts can be cut or ended for convenience, so obligations called non-cancelable remain exposed to budgets.
How we value it
No fair value: Infleqtion loses money at the operating line every quarter and analysts forecast losses through 2029, so a value built on earnings would rest on a guess; the bet prices it instead.
How we read the multiple
With no profit to price, we split the price between the two businesses.
- The whole company, netting its cash, is worth about $1.26B.
- Sensors and software bring about two thirds of the $45.1M expected this year, about $30M.
- At 4 times sales, generous for work at a loss, that is about $120M.
- That leaves about $1.1B for the computer.
- The computer sells about $15M a year, and no outside judge has picked it.
Against its own history. It listed in February 2026, so there is no earlier history; buyers pay far less for each dollar of sales than at the April high.
The evidence under this - 4 claims
The vendor's bridge reads the $640.4M of total assets as preferred stock and reaches about $1,959M; the filed balance sheet shows no preferred stock.
Pure-play rivals trade at hundreds of times sales, so the market may value a computing roadmap that is on schedule far above any multiple of today's sales.
Each rival's revenue is under about $30M a quarter except IonQ's, so these multiples are struck on small bases and price the theme.
The 10-Q reports one segment and no split by product line, so the mix cannot be checked.
Priced against its rivals
The competitors its own annual report names, each confirmed against the rival's filing. Measured on company value for each $1 of sales, cheapest first:
| Ticker | Company | company value for each $1 of sales |
|---|---|---|
| IONQ | IonQ | 26.2× |
| INFQ | Infleqtion | 62.7× |
| RGTI | Rigetti Computing | 204.1× |
| QBTS | D-Wave Quantum | 213.5× |
| QNT | Quantinuum | 304.4× |
| IBM | International Business Machines not in the median | 4.1× |
The middle of the other 4 is 208.8. International Business Machines Corporation competes with it in one product line inside a much bigger company, so its price says little about this business and it is not counted in the middle. Infleqtion Inc is 2nd cheapest of 5.
Rivals' figures come from a data provider, so read the spread of the group rather than any single pair.
The group's pure plays cost far more for each dollar of sales; that is the whole theme's excitement, not a sign Infleqtion is cheap.
Also checked and set aside: Microchip Technology Incorporated (a substitute, not the same product).
Private or foreign rivals that cannot be priced: PsiQuantum, QuEra Computing, Atom Computing, Safran Electronics & Defense, Oscilloquartz.
The bet: what it pays if it works
This bet is on the quantum computer. If it reached $100M of sales a year, more than analysts expect for the whole company in 2029, and buyers paid 20 times sales for it, the company would be worth less than twice today's value.
If the computer never wins large orders, the sensor work and the cash would be worth about a third of today's price, less years of losses.
| Quantum computing: Sqale systems and cloud access revenue in two to three years | Million dollars of computing revenue a year needed | At the stated multiple |
|---|---|---|
| $40.00M | about 40 at 100% kept | 0.7× |
| $100.00M (the big win) | about 100 at 100% kept | 1.7× |
| $200.00M | about 200 at 100% kept | 3.3× |
To pay 2× (a Fair bet) it would need about $119.57M of Quantum computing: Sqale systems and cloud access revenue (about 120 million dollars of computing revenue a year); to pay 5× (a Cheap bet), about $307.91M of Quantum computing: Sqale systems and cloud access revenue (about 308 million dollars of computing revenue a year), at the same multiple.
Each row is Quantum computing: Sqale systems and cloud access alone, plus the rest of the company (Quantum sensing, timing and software contracts) counted at about $120.00M.
The steps it has to take
- Done 12 logical qubits shown
- Done Two quantum computers delivered, to the UK and Japan
- Done 30 logical qubits shown, the 2026 target
- Not started Selection for a DARPA benchmarking stage due 2027-12
- In progress 100 logical qubits due 2028-12
- Not started Computing above half of revenue due 2029-12
The detail behind the label
The evidence under this - 4 claims
No outside body has verified the result, and logical-qubit counts are not measured the same way across companies.
Pure-play rivals trade at hundreds of times sales, so the market may value a computing roadmap that is on schedule far above any multiple of today's sales.
The company attributes part of the gross loss to stock pay and the timing of a NASA contract's costs, so the gross loss is partly not recurring.
Where we disagree with the market
The cheapest atom-based quantum company, with cash and a computer on schedule.
A government contractor whose computer is priced at a billion dollars with no outside endorsement.
The market sees the cheapest listed company using atoms for quantum computing, after a 63% fall, with years of cash. We see a contractor whose sales rest on one government customer, with a computer an outside judge has not chosen, still priced at over a billion dollars for it.
The evidence against us: a DARPA selection, and sales spreading beyond one customer, would show the computer earning its price.
What analysts say, and the evidence
The evidence under this - 5 claims
Three analysts make the 2026 figure, and the share of buys fell from 100% of five in July.
Each rival's revenue is under about $30M a quarter except IonQ's, so these multiples are struck on small bases and price the theme.
Pure-play rivals trade at hundreds of times sales, so the market may value a computing roadmap that is on schedule far above any multiple of today's sales.
DARPA opened a new Stage A solicitation due 30 November 2026, so a later entry remains possible.
A government customer funding several programmes is harder to lose at once than a commercial buyer, though the 10-Q does not name it.
What could go wrong, and how we would know
The risks
- One customer. 81% of third-quarter sales came from one government customer.
- Weak accounts. Sales were revised for three periods, a filing came late, and controls are still weak.
- Passed over. DARPA picked two rivals using the same atoms, not Infleqtion.
- Pay in shares. Stock pay about equal to sales keeps adding shares.
The evidence under this - 5 claims
A government customer funding several programmes is harder to lose at once than a commercial buyer, though the 10-Q does not name it.
The company judged the adjustments not material to the earlier periods.
DARPA opened a new Stage A solicitation due 30 November 2026, so a later entry remains possible.
Much of 2026's stock pay is awards that vested when the company listed, so the quarterly rate may fall.
What we still do not know
- Who the main customer is, and how many separate programmes it funds.
- How much of sales really comes from the computer, which the filings do not split out.
- Whether the 30-qubit result holds up to outside checks.
The evidence under this - 3 claims
A government customer funding several programmes is harder to lose at once than a commercial buyer, though the 10-Q does not name it.
The 10-Q reports one segment and no split by product line, so the mix cannot be checked.
No outside body has verified the result, and logical-qubit counts are not measured the same way across companies.
What would change our view
- Unattractive→NeutralAn outside judge picks the computer
It needs all of:
- ○DARPA selects Infleqtion for a stage of its quantum computer programme.watched as it happens
- ○The whole company's value falls to 20 times its last year's sales.about ≈$5.39 today, 32% below the last close
- Unattractive→NeutralA price that pays little for computing
It needs:
- ○The whole company's value falls to 10 times its last year's sales.about ≈$2.70 today, 66% below the last close
We would also be wrong if
- Fourth-quarter 2026 sales finish below $12.5M, missing the year's outlook.
- Another accounting correction or late filing in 2027.
- One customer stays above half of sales through 2027.
What is coming
- 30 November 2026 DARPA Stage A solicitation closes: Whether Infleqtion bids again for the benchmarking programme that picked its rivals.
Every dated event ahead
- 30 November 2026 DARPA Stage A solicitation closes: Whether Infleqtion bids again for the benchmarking programme that picked its rivals.
- 30 March 2027 Q4 2026 results and 2027 outlook: Whether 2026 reached $45.1M and how much of 2027 rests on Customer A.
- 31 December 2028 100 logical qubits: Whether the computer reaches the point management says starts computing growth.
Since we wrote this
- 17 NovThesis writtenUnattractive, medium convictionWritten the day after the Q3 2026 results, with the shares 63% below their April high.
Sources 14
Sources - 14
Every claim above cites one of these, or states the reasoning that derived it. The quote is frozen at the date it was read: a link can rot, and an agent fetch returns a reading of a page rather than the page. Backs counts the whole claim ledger, so it can name a claim this page does not surface.
- S01primarycompany filing2026-03-31the business, the products, Sqale's atom count, logical qubits and gate fidelity, the named competitors, the government awards and the risk factorsbacksC01C19C22C29
- S02primarycompany filing2026-11-16revenue by type and by country, customer concentration, government share, cash and securities, cash flows, stock pay, remaining performance obligations, share count, options and warrants, the material weaknesses and the liquidity statementbacksC02C03C04C05C06C07C08C09C10C11C12C13C14C17C18C31C33C34C41
- S03primarycompany release2026-11-16the quarter's revenue, gross loss, operating loss on both bases, the maintained 2026 outlook, the 30-logical-qubit result, the NASA expansion and the direction-finding awardbacksC01C02C06C09C15C20C28C30C32
- S04primarycompany release2026-04-08Legacy Infleqtion's 2025 revenue as first reported, operating loss on both bases, stock pay and cashbacksC03C07C08
- S05primarytranscript2026-04-08the 2026 revenue guide of about $40M, the sensing and computing split, the two systems delivered, the Safran partnership and the logical-qubit roadmap
Matt mentioned 2/3, 1/3 that continue to be the trend. And that's until we reach the point that Pranav mentioned in 2028 when we reach 100 [indiscernible] logical qubits
backsC15C20C21C29C30 - S06primarycompany filing2026-04-10the officers and their history, the beneficial ownership table, 2025 executive pay and the bonus terms, the business combination proceeds and the warrantsbacksC13C38C39C40
- S07primarygovernment data2026-10-08the companies in Stages B and C, including Atom Computing in Stage C and QuEra in Stage B, the Stage C award size of up to $300M, and the absence of Infleqtion from both lists
Unlike classical computing, no single dominant architecture exists.
backsC23C24 - S08primarygovernment data2024-12-01US federal spending on quantum research and development by year
Agencies reported budget expenditures for QIS R&D of $456 million in FY 2019, $690 million in FY 2020, $851 million in FY 2021, $1,041 million in FY 2022, and $1,036 million in FY 2023, followed by $1,006 million of enacted budget authority for FY 2024 and a requested budget authority of $998 million for FY 2025.
backsC25 - S09primarycompany filing2026-02-25that IonQ bought its way into quantum sensing and optical clocks, including Vector Atomic
We also cemented our leadership in quantum networking, quantum sensing and quantum security with our acquisitions of id Quantique SA
backsC26 - S10primarycompany filing2026-03-04that a rival sells cloud access and systems to the same government and research buyers
We provide access on a commercial basis to our quantum computers over QCS, directly engaging with enterprises and government organizations making significant investments in quantum computing research, development, and readiness.
backsC27 - S11secondaryvendor data2026-11-17Research pack (FMP market tier and FINRA short interest), frozen 2026-11-17content/INFQ/research-pack.jsonthe close, the 205-session high and low, returns, the analyst split and estimates, the Q3 consensus, short interest and the rivals' sales multiplesbacksC31C32C33C34C36C44
- S12primarycompany filing2026-08-21each open-market sale and purchase, its price and whether it was made under a pre-set planbacksC37
- S13primarycompany filing2026-08-17the revised quarterly revenue for 2025 and the first quarter of 2026, and the second quarter's revenuebacksC02C03
- S14primarycompany filing2026-08-17the revenue recognition error, the periods revised, and its tie to the material weaknessbacksC16